From: Jerry Sievers <gsievers19@comcast.net>
To: pgsql-admin@postgresql.org
Subject: using pg_receivewal --drop-slot as recovery.conf:recovery_end_command
Date: Wed, 19 Dec 2018 14:28:27 -0600
Message-ID: <87h8f945xw.fsf@jsievers.enova.com> (raw)
Admins, I tested $subject and it works like a charm for dropping a no
longer needed slot on physical replication provider during promote of
the standby.
Wondering though about possible race condition...
I see in server logs on standby side that we sigterm the walreceiver but
speculate about cases where the provider side (for any reason) is
delayed in noticing the receiver death...
In turn leading to a refusal to allow slot dropping.
Much ado a about nothing?
pg_receivewal --slot $foo --drop-slot --dbname 'same stuff as primary
conninfo'
This was compelling to use since we are preparing to promote some
standbys at the stroke of midnight inclusive=false for year end
reporting and will *not* be using our external standby0-promote tool
which already knows how to drop the provider slot.
Please advise.
Thx
--
Jerry Sievers
Postgres DBA/Development Consulting
e: postgres.consulting@comcast.net
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Reply to all the recipients using the --to and --cc options:
reply via email
To: pgsql-admin@postgresql.org
Cc: gsievers19@comcast.net
Subject: Re: using pg_receivewal --drop-slot as recovery.conf:recovery_end_command
In-Reply-To: <87h8f945xw.fsf@jsievers.enova.com>
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
This inbox is served by DDX for PostgreSQL; see mirroring instructions
for how to clone and mirror all data and code used for this inbox